Southern Minnesota pork producer Robert Baarsch is looking for a bigger piece of the packer pie. To get it, the CEO of Next Generation Pork has been working hard to eliminate variation in his pigs.

There is easily $20 a pig on the packing side if we could drive out variation,” says the LeRoy, MN, producer.

Baarsch has been using Farmweld automatic sorting technology (FAST) since its inception several years ago. A tighter sort in his finishing barns has resulted in a 10%-plus increase in the number of carcasses falling in the “red box” — the spot on the Hormel matrix where bonus dollars are earned beyond base price.

But it doesn't stop there. He is studying pig behavior to discover why, when and where there is variation.

“We knew the scale would sort pigs better, but we were measuring averages of groups,” Baarsch points out. “We're learning that individual pig behavior may be a key variable causing variation.”

The next generation of sorting technology — the FAST II system — allows Baarsch and his managers to monitor individual pigs. The scales carry the HerdStar electronics his team developed for radio frequency identification (RFID). Antennas and RFID readers are installed on the scales to gather data on every pig wearing a RFID eartag. Not only can they sort better, but, Baarsch says, they'll “nip variation in the bud.”

HerdStar is a technology company Baarsch founded almost three years ago to provide near real-time data for grow-finish management. The RFID and electronic components for the FAST II scale are sold through its Web site at

When Baarsch first tested the RFID system over a year ago, several observations surprisingly stood out:

  • Growth rate for individual pigs was very linear. So linear, in fact, that it was easy to see even a 51-hour hiccup in growth in a particular pig. The interruption happened at about the same time in other tagged pigs. The reason for the slight blip in growth rate hasn't been determined, but Baarsch believes it marks the beginning of a new way to study pig behavior and performance on an individual level.

  • Once an animal is set on a growth rate, it didn't change much during the finishing phase. In fact, they observed, 80% of the variation in rate of gain can be explained by pig-to-pig differences vs. variability within individuals.

  • Pig eating patterns are predictable, with preferred times of the day. Visits to the food court increased in the morning between 8:00 and 9:00, with the biggest activity spike around 6:00 p.m.

  • Pigs are lighter in the morning and grow through the day.

  • After the first sort for marketing, growth rate of the remaining pigs in the barn was unchanged.

  • Lighter pigs at weaning were the lighter pigs at market. In other words, the 40-lb. pig never ended up heavier than the 50-lb. pig. “That tells me variation on end weight goes all the way back to the farrowing house,” notes Baarsch.

For Research Purposes Only

For now, Baarsch is only using RFID for research purposes. He calls it trackability rather than traceability.

“Traceability infers food safety and RFID can do that. But we're hot on RFID because it tells us which pigs are growing faster, mortality rates and health events,” he says. “We can see a downtick in growth within 24-36 hours. If a pig has a flutter, we can pick it up. Retinal scanning and DNA as identification methods don't do anything for a producer except add cost and liability.”

Costs of the technology will likely drop in time, but at today's price of $2.50/tag and $2,500 for a reader, RFID won't reach every pig.

“But large systems will want a few of these barns with this capability,” Baarsch states. “As university research dollars dry up, farms will need to do their own research to make buying decisions.”

The eartags were developed after much trial and error by Digital Angel, a South St. Paul, MN, company. At first, retention level was poor, according to Baarsch, who was involved in their research and development. But the back of the tag was changed and the tag was placed higher on the ear so retention is a respectable 98%. That's not bad considering tags go on at 3 to 5 days of age, and stay on for five months in an environment that's probably one of the worst for electronics.

Tags must also survive scalding hot water at the packing plant, scorching, blast chilling and still retain readability, if carcass data is to be collected.

Database Ties Into Scale

Only about 5% of Next Generation Pork's 100,000 pigs are tagged with RFID, but they provide a lot of information. One 600-pig, wean-to-finish barn will generate 1,500 records if pigs cross the scale 2.5 times a day. “That many records would overload a laptop in 30 days,” states Baarsch.

Not to worry. Baarsch and his software team developed a Web-interfaced database called eBarn for FAST II customers to collect the wealth of data. Information such as temperature variation, scale activity, water usage, rate of gain and mortalities are put into 5- and 30-day reports through eBarn's grow-finish management software.

Baarsch plans to launch a HerdStar program that “paints by RFID” at the winter trade shows.

He explains: A pig steps on a scale that is Internet-enabled. A computer detects that the pig hasn't grown for a second day in a row, so statistically, something is amiss. A paint sprayer is activated to flag the pig and an e-mail is sent to the barn alerting the manager to check it.

“Not many people are trying to recover money left on the table for marketing,” Baarsch says. “Packers like Hormel are going for exact portion sizing in packaging. They're also trying to robotize processing so primals have to be kept within certain parameters.

“We need to control the size of the loin because they want everything the same. Different size pork chops will never let us compete with chicken,” he adds.

“We just need to look at pigs as individuals. That's the concept to driving out variation. RFID is a way to unleash the power of the individual pig and to solve problems on an individual basis. We have to adopt technology and manage it. We can't settle for average, it's not good enough,” Baarsch concludes.